Universal Mobile Telecommunications System (UMTS) is an example of a mobile radio communications system. UMTS is a 3rd Generation (3G) mobile communication system employing Wideband Code Division Multiple Access (WCDMA) technology standardized within the 3rd Generation Partnership Project (3GPP). In the 3GPP release 99, the radio network controller (RNC) in the radio access network controls radio resources and user mobility. Resource control includes admission control, congestion control, and channel switching which corresponds to changing the data rate of a connection. Base stations, called node Bs (NBs), which are connected to an RNC, orchestrate radio communications with mobile radio stations over an air interface. RNCs are also connected to nodes in a core network, i.e., Serving GPRS Support Node (SGSN), Gateway GPRS Support Node (GGSN), mobile switching center (MSC), etc. Core network nodes provide various services to mobile radio users who are connected by the radio access network such as authentication, call routing, charging, service invocation, and access to other networks like the Internet, public switched telephone network (PSTN), Integrated Services Digital Network (ISDN), etc.
The Long Term Evolution (LTE) of UMTS is under development by the 3rd Generation Partnership Project (3GPP) which standardizes UMTS. There are many technical specifications hosted at the 3GPP website relating to Evolved Universal Terrestrial Radio Access (E-UTRA) and Evolved Universal Terrestrial Radio Access Network (E-UTRAN), e.g., 3GPP TS 36.300. The objective of the LTE standardization work is to develop a framework for the evolution of the 3GPP radio-access technology towards a high-data-rate, low-latency and packet-optimized radio-access technology. In particular, LTE aims to support services provided from the packet switched (PS)-domain. A key goal of the 3GPP LTE technology is to enable high-speed packet communications at or above about 100 Mbps.
FIG. 1 illustrates an example of an LTE type mobile communications system 10. An E-UTRAN 12 includes E-UTRAN NodeBs (eNBs) 18 that provide E-UTRA user plane and control plane protocol terminations towards the user equipment (UE) terminals 20 over a radio interface. An eNB is sometimes more generally referred to as a base station, and a UE is sometimes referred to as a mobile radio terminal or a mobile station. As shown in FIG. 1, the base stations are interconnected with each other by an X2 interface. The base stations are also connected by an S1 interface to an Evolved Packet Core (EPC) 14 which includes a Mobility Management Entity (MME) and to a System Architecture Evolution (SAE) Gateway. The MME/SAE Gateway is shown as a single node 22 in this example and is analogous in many ways to an SGSN/GGSN gateway in UMTS and in GSM/EDGE. The S1 interface supports a many-to-many relation between MMEs/SAE Gateways and eNBs. The E-UTRAN 12 and EPC 14 together form a Public Land Mobile Network (PLMN). The MMEs/SAE Gateways 22 are connected to directly or indirectly to the Internet 16 and to other networks.
To provide efficient resource usage, LTE and other systems that used shared radio resources support fast “dynamic” scheduling where resources on the shared channels, e.g., in LTE this includes the physical downlink shared channel (PDSCH) and the physical uplink shared channel (PUSCH), are assigned dynamically to user equipment (UE) terminals and radio bearers on a sub-frame basis according to the momentary traffic demand, quality of service (QoS) requirements, and estimated channel quality. This assignment or scheduling task is typically performed by one or more schedulers situated in the eNB.
The overall scheduling concept for the downlink is illustrated in FIG. 2. To support fast channel-dependent link adaptation and fast channel-dependent time and frequency domain scheduling, the UE 20 may be configured to report the Channel Quality Indicator (CQI) to aid the eNB 18 in its dynamic scheduling decisions. Typically, the UE 18 bases the CQI reports on measurements on downlink (DL) reference signals. Based on the CQI reports and QoS requirements of the different logical channels, the DL scheduler in the eNB 18 dynamically assigns time and frequency radio resources, i.e., scheduling blocks. The dynamically-scheduled radio resource assignment is signaled on the Physical Downlink Control Channel (PDCCH) in the LTE example. Each UE 20 monitors the control channel to determine if that UE is scheduled on the shared channel (PDSCH in LTE), and if so, what physical layer radio resources to find the data scheduled for downlink transmission.
The uplink scheduling concept is illustrated in FIG. 3. The UE 20 informs the UL scheduler in the eNB 18 when data arrives in the transmit buffer with a Scheduling Request (SR). The UL scheduler selects the time/frequency radio resources the UE will use and also selects the transport block size, modulation, and coding because link adaptation for the uplink is performed in the eNB. The selected transport format is signaled together with information on the user ID to the UE. This means that the UE must use a certain transport format and that the eNB is already aware of the transmission parameters when detecting the UL data transmission from that UE. The assigned radio resources and transmission parameters are sent to the UE via the PDCCH in LTE. Later, additional Scheduling Information (SI) such as a Buffer Status Report (BSR) or a power headroom report may be transmitted together with data.
Although dynamic scheduling is the baseline for LTE and other systems, it can be less than optimum for certain types of services. For example, for services such as speech (VoIP) where small packets are generated regularly, dynamic scheduling results in substantial control signaling demands because a radio resource assignment needs to be signaled in each scheduling instance, which in the case of VoIP, an assignment must be signaled for every VoIP packet. To avoid this high relative signaling overhead for these types of services, resources may be assigned semi-statically, which is called “semi-persistent” or “persistent” scheduling. A semi-persistent assignment is only signaled once and is then available for the UE at regular periodic intervals without further assignment signaling.
Many modern wireless communications systems use a hybrid ARQ (HARQ) protocol with multiple stop-and-wait HARQ “processes”. The motivation for using multiple processes is to allow continuous transmission, which cannot be achieved with a single stop-and-wait protocol, while at the same time having some of the simplicity of a stop-and-wait protocol. Each HARQ process corresponds to one stop-and-wait protocol. By using a sufficient number of parallel HARQ processes, a continuous transmission may be achieved.
FIG. 4 shows an eNB 18 with an HARQ controller 22 that includes multiple HARQ entities 1, 2, . . . , m (24), with each HARQ entity managing HARQ processes for a corresponding active UE 1, 2, . . . , n (20). FIG. 5 shows each HARQ entity 24 managing one or more HARQ processes A, B, . . . , n (26). One way of looking at the HARQ process is to view it as a buffer. Each time a new transmission is made in an HARQ process, that buffer is cleared, and the transmitted data unit is stored in the buffer. For each retransmission of that same data unit, the received retransmitted data unit is soft-combined with the data already in the buffer.
FIG. 6 illustrates an example of the HARQ protocol where P(X,Y) refers to the Yth transmission in HARQ process X. The example assumes six HARQ processes. If a large number of higher layer packets (e.g. IP packets) are to be transmitted, for each transmission time interval (TTI), the RLC and MAC protocol layers perform segmentation and/or concatenation of a number of packets such that the payload fits the amount of data that can be transmitted in a given TTI. The example assumes for simplicity that one IP packet fits into a TTI when RLC and MAC headers have been added so that there is no segmentation or concatenation.
Packets 1 through 6 can be transmitted in the first six TTIs in HARQ processes 1 through 6. After that time, HARQ feedback for HARQ process 1 is received in the receiver. In this example, a negative acknowledgment (NACK) for HARQ process 1 is received, and a retransmission is performed in HARQ process 1 (denoted P1,2). If a positive acknowledgment (ACK) had been received, a new transmission could have started carrying packet 7. If all 6 first transmissions failed (i.e., only NACKs are received), then no new data can be transmitted because all HARQ processes are occupied with retransmissions. Once an ACK is received for an HARQ process, new data can be transmitted in that HARQ process. If only ACKs are received (no transmission errors), then the transmitter can continuously transmit new packets.
In modern cellular systems, synchronous HARQ may be used for the uplink and asynchronous HARQ for the downlink. For that case, in the uplink, the subframe or transmission time interval (TTI) when the retransmission occurs is known at the base station receiver, while for the downlink, the base station scheduler has the freedom to choose the subframe or TTI for the retransmission dynamically. For both uplink and downlink, a single-bit HARQ feedback (ACK/NACK) is sent providing feedback about the success of the previous data unit transmission.
A problem created by introducing semi-persistent scheduling, as is currently proposed for LTE for example, is that a receiving UE cannot match-up a dynamically-scheduled retransmission of a HARQ process with the initially-transmitted HARQ process that was semi-persistently scheduled. If HARQ is operated in asynchronous mode, as is currently proposed for example in the LTE downlink, the problem is how the HARQ processes should be selected for semi-persistent scheduling. After a semi-persistent assignment, both the HARQ transmitter entity as well as the HARQ receiver entity would, for example, randomly pick an idle HARQ process with potentially different HARQ process IDs. The reason is that the eNB does not send an explicit assignment referring to a particular HARQ process ID. If the HARQ receiver can decode the information, it delivers the information to higher layers and acknowledges the reception. But if decoding fails, tehn the HARQ receiver sends a negative acknowledgement, and the HARQ transmitter issues a retransmission of that HARQ process. If the retransmission is scheduled dynamically (as in the LTE downlink), then the corresponding dynamic assignment must contain the identifier of the HARQ process. It is likely that the HARQ transmitter chose a HARQ process ID for the initial transmission that was different from the HARQ process ID selected by the HARQ receiver. Consequently, the HARQ receiver cannot match the dynamically retransmitted HARQ process unambiguously to a pending HARQ process. In fact, there may be multiple pending processes (persistently or dynamically scheduled) for which the receiver may not even have received the assignment. If different HARQ processes are used by the transmitter and the receiver, then the data may be erroneously soft-combined with other data and the transmitted can not correctly identify the HARQ ACK/NACK sent for the data. The failure to make this match thus significantly increases error rate and decreases throughput.